One of the approaches in the immunotherapy of breast cancer patients involves vaccination with peptides derived from tumor-associated antigens specifically designed to associate with T cells in the context of major histocompatibility complex (MHC) class I and class II molecules. This approach is non-toxic and has the potential of controlling disease and prolonging time to recurrence and ultimately, even serving as a preventive measure. Tumor specific antigens, including MUC1 and tumor associated carbohydrate antigens (TACA) have been shown to elicit tumor-specific cytotoxic T lymphocytes (CTLs) and establish a non-self reactive long-term immune memory. MUC1 is a widely expressed tumor- antigen that is over expressed and aberrantly glycosylated on greater than 90% of breast carcinomas and has been shown to elicit tumorspecific immunity. Several clinical trials using non-glycosylated MUC1 tandem repeat sequences have elicited MUC1-specific CTLs with limited effects on clinical response. However, MUC1 occurs naturally as a heavily glycosylated protein which contains TACA. TACA-containing glycopeptides are appealing CTL-based vaccines as they are widely expressed in breast tumors, their expression is largely tumor specific, and they are induced early during neoplastic transformation. Our objective is to develop optimal breast cancer-based vaccines comparing glycosylated and non-glycosylated natural and anchor-improved peptides derived from the MUC1 sequence. Our hypothesis is that eliciting immunity to glycosylated and/or anchor-improved MUC1 peptides will result in anti-tumor immunity and long-term immune memory. The vaccine design addresses breast cancer prevention as well as immunotherapy after tumor onset. Our specific aims are 1) to generate and characterize CTLs specific for Kb and Db MHC class l-binding MUC1 peptides and assess their effectiveness in the prevention and treatment of spontaneous mammary gland carcinomas in mouse models, 2) to determine in vitro the optimal MUC1 peptides for immunotherapy in HILA-A*0201 individuals, 3) to ascertain "immune competence" of breast cancer patients following primary cancer therapy, and 4) to translate the most effective vaccine strategies in phase I clinical trials in patients with minimal residual disease.